Short-range and long-range context effects on coliphage T4 endonuclease II-dependent restriction

Abstract

Synthetic sites inserted into a plasmid were used to analyze the sequence requirements for in vivo DNA cleavage dependent on bacteriophage T4 endonuclease II. A 16-bp variable sequence surrounding the cleavage site was sufficient for cleavage, although context both within and around this sequence influenced cleavage efficiency. The most efficiently cleaved sites matched the sequence CGRCCGCNTTGGCNGC, in which the strongly conserved bases to the left were essential for cleavage. The less-conserved bases in the center and in the right half determined cleavage efficiency in a manner not directly correlated with the apparent base preference at each position; a sequence carrying, in each of the 16 positions, the base most preferred in natural sites in pBR322 was cleaved infrequently. This, along with the effects of substitutions at one or two of the less-conserved positions, suggests that several combinations of bases can fulfill the requirements for recognition of the right part of this sequence. The replacements that improve cleavage frequency are predicted to influence helical twist and roll, suggesting that recognition of sequence-dependent DNA structure and recognition of specific bases are both important. Upon introduction of a synthetic site, cleavage at natural sites within 800 to 1,500 bp from the synthetic site was significantly reduced. This suggests that the enzyme may engage more DNA than its cleavage site and cleaves the best site within this region. Cleavage frequency at sites which do not conform closely to the consensus is, therefore, highly context dependent. Models and possible biological implications of these findings are discussed.